logprior.distance: Log Prior Distance

In anthony-aylward/asepirinen: Assessing allele-specific expression across multiple tissues from RNA-seq read data

Description

Counts the prior probability for each heterogeneous state as a function of its distance.

Usage

logprior.distance(m, p0 = 0.75, p.dist = NULL,
  model.strong.ase = TRUE)

Arguments

m	the number of tissues
p0	the joint prior probability of the 3 homogeneous states
p.dist	either a vector of length 'max.dist' of total probabilities of each set of states for distances 1,...,'max.dist' Where max.dist==m-ceiling(m/3) if model.strong.ase==TRUE and max.dist==floor(m/2) if model.strong.ase==FALSE. Interpreted as relative to each other so that after renormalisation and scaling sum to (1-p0). OR, if p.dist == NULL, then p.dist will be set uniform =(1-p0)/max.dist over the distance.

Details

Distance is the smallest number of changes that turns the state into one of the homogeneous states if model.strong.ase==TRUE then the maximum distance is m-ceiling(m/3), if model.strong.ase==FALSE then the maximum distance is floor(m/2), minimum distance is 1 for a heterogeneous configuration (the three homogeneous configurations have distance 0)

Value

log.prior

is the log of prior probability of each STATE as a function of distance 1,...,max.dist results from dividing p.dist by the number of states in each distance category

log.sum.prior.h0

is the log of the sum of priors over all heterog states that have at least one 0

log.sum.prior.h1

is the log of the sum of priors over all heterog states that have no 0

anthony-aylward/asepirinen documentation built on May 13, 2019, 11:29 a.m.